Over time, new words enter the language, others become obsolete, and existing words acquire new meanings. These phenomena are grounded in a fascinatingly complex mix of cognitive, social, and contextual factors, responding to language contact, emerging circumstances, cultural and socio-political changes, stylistic choices, and different communicative needs. The past decade has seen a growing interest in automatic methods for semantic change (i.e. meaning change) detection from large corpus data, which have made it possible to conduct quantitative studies aimed at detecting broad patterns in the data. Most of these automatic detection methods rely on distributional semantics methods and trace the computational representation of a corpus-driven word’s semantic profile (via vector embeddings) over time to identify if and when a potential change in the semantic profile may have occurred. In this talk I will present my research on developing computational models for semantic change detection in historical texts, particularly on ancient Greek and Latin.
McGillivray, B. et al. (2019). A computational approach to lexical polysemy in Ancient Greek, Digital Scholarship in the Humanities, 34: 4.
McGillivray, B. et al. (2022). A new corpus annotation framework for Latin diachronic lexical semantics. J. Latin Linguistics, vol. 21, No. 1, pp. 47-105.
Talk will be broadcasted at:
https://us06web.zoom.us/j/88466865139?pwd=VnFwZXBxZUdHRUpUVkJTM2FVSXc2Zz09
